Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
The best programmers are not necessarily algorithm experts but deliver results and have a lot of curiosity. They prioritize building a product for the user, choose the right tools for the job regardless of prior experience, and write simple, easily adaptable code.
Scalar is an open-source API platform with a modern REST API client, API references, and first-class OpenAPI/Swagger support.
Tiptap UI Components is a library of React components, templates, and primitives for building rich text editors easily on top of the Tiptap editor framework.
FumaDocs is a documentation framework built on Next.js that uses TypeScript, MDX, and Tailwind CSS. It provides utilities, a UI, and features of Next.js App Router. FumaDocs includes built-in components and integrations, focusing on a user-friendly mobile responsive experience.
A customer's UI thread was hanging due to a deadlock involving a watchdog thread. The watchdog thread was suspending the UI thread to capture a stack trace, but the UI thread held a lock needed by the watchdog to complete the stack capture, resulting in a deadlock. To avoid this, thread suspension for debugging should be done from a separate process.
PgDog is a network proxy that understands the Postgres wire protocol, allowing it to route queries to multiple databases without application code changes. It parses SQL queries using a Rust library extracted from Postgres itself to identify sharding keys and determine where to send the query. It also manages cross-shard queries by handling and manipulating messages to combine results from multiple shards, including rewriting row counts and managing data types.
Notion Mail is an AI-powered inbox that organizes, drafts, and schedules emails with features like auto-filtering, custom views, and built-in scheduling.
Google One AI Premium subscribers can now generate and share videos from text prompts in Gemini Advanced and animate images in Whisk using the Veo 2 model.
MCPify is a no-code tool for building MCP servers that can be used by Claude, Cursor, and other AI tools.
protobuf-ts-types is a TypeScript library that infers types from protobuf message definitions without code generation or compilation using TypeScript's template literal types.
Sponsored
Bugs sneak out when less than 80% of user flows are tested before shipping. But getting that kind of coverage—and staying there—is hard and pricey for any size team.
Apple will start training artificial intelligence models by analyzing data on customers' devices to improve its AI platform. The company currently trains its models using synthetic data, but synthetic information isn't always representative of actual customer data. Customer data will remain on-device and won't be directly used to train AI models. The new system will roll out in an upcoming beta of iOS and iPadOS 18.5 and macOS 15.5.
The FTC's antitrust case against Meta seeks to compel the company to undo its acquisitions of Instagram and WhatsApp. Meta allegedly wields an illegal monopoly in social media. The company says that the FTC's case ignores how people use technology today - Meta currently faces stiff competition from YouTube, TikTok, and other video platforms. It plans to demonstrate that competition is alive and well in the social networking industry.
A new AI model from Google running on Pixel phones is helping researchers better understand the language of dolphins. DolphinGemma was created using dolphin language data collected over 40 years. It was built specifically for Pixel devices. Google plans to share DolphinGemma as an open model this summer. The research may eventually enable us to construct sounds that dolphins might understand.
German researchers have revived algae cells that were dormant for over 7,000 years. Found buried at the bottom of the Baltic Sea imprisoned under layers of sediment, the cells were deprived of oxygen or light for millennia. They showed full function recovery after being revived and started multiplying again with no issues. It is now possible to conduct 'time-jump experiments' into the various stages of Baltic Sea development in the lab.
gh-signoff is a GitHub command-line extension for local CI. It enables developers to run CI on their own machines and sign off when they pass. Developer machines are super-fast nowadays and chronically underutilized - Cloud CI services are typically slow, expensive, and rented. gh-signoff brings CI back in-house.
The job market is currently having a tough time, but this isn't because companies are actually replacing developers with autonomous AI agents. There are macroeconomic factors, layoffs, and other things leaving tons of highly qualified devs out there looking for work. Some companies are operating under the myth that AI will make developers obsolete soon so they're not hiring as aggressively as they otherwise would. However, despite AI evangelists claiming that, any day now, companies won't need human developers to build their products, human developers are still needed.
French bailiffs hired by Rippling were unable to locate Deel's CEO, Alex Bouaziz, during attempts to serve him with papers as part of a lawsuit alleging Deep paid one of its employees to spy on Rippling. Bouaziz is known to be 'highly nomadic' but remains very attached to France, where he has a listed address. Rippling may need to apply to serve him via email.
OpenAI has launched a new family of models called GPT-4.1 that excels at coding and instruction-following. The models, GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano, are available through OpenAI's API but not ChatGPT. They are multimodal and have a 1-million-token context window. GPT-4.1 costs $2 per million input tokens and $8 per million output tokens, GPT-4.1 mini is $0.40/million input tokens and $1.60/million output tokens, and GPT-4.1 nano is $0.10/million input tokens and $0.40/million output tokens.
Sponsored
Spreadsheets are awesome. AI makes them even better. Chat with your data, chart key metrics with built-in Python, and auto-generate SQL
Privacy is less needed for people whose life situations are relatively normal, and more needed for people whose life situations deviate from the norm, in any direction - and you never know when you will become someone who needs it.
Interoperable apps enable users to easily explore different experiences or their data, allowing applications to compete on an experience of users' data rather than on owning it.
OpenAI may soon need government ID verification for organizations to access certain models - not all organizations will be eligible for verification.
Improving bad writing can be more effort than writing something well, and large language models make too many mistakes to be worth using.
MCP has rapidly grown into the de-facto standard for integrating third-party data and tools with AI agents, but it has a lot of nuanced vulnerabilities and limitations.
The deal values Altera at $8.75 billion and will make the division operationally independent - Intel will retain a 49% stake in Altera.
Sponsored
Delve officially launched Computer Use Agents that allow founders and GRC teams to go auto-capture all screenshots for SOC 2.
OpenAI has launched three new models in its API: GPT‑4.1, GPT‑4.1 mini, and GPT‑4.1 nano. These models outperform GPT‑4o and GPT‑4o mini across the board, with major gains in coding and instruction following. They also have larger context windows—supporting up to 1 million tokens of context—and are able to better use that context with improved long-context comprehension. They feature a refreshed knowledge cutoff of June 2024.
Hugging Face, the center of the open source AI community, has long stated its goal is to be a decentralized DeepMind. While this isn't exactly the case, adding in an open source robotics platform via Pollen moves it closer to that goal.
DeepMind has announced DolphinGemma, a large language model developed by Google that helps scientists study how dolphins communicate — and hopefully find out what they're saying, too.
The ByteDance team has released a paper showing how to train a competitive 7B parameter video generation model on a “modest” compute budget of 655k H100 hours. It has strong performance on a number of temporally difficult tasks.
Most generative models on continuous signals operate in latent space due to computational constraints. This work introduces a series of cascades that allow the generation to happen directly in pixel space. This eliminates the need for a pretrained VAE.
New VLM that can reason about contacts between humans in 3D and objects. It does so by leveraging a strong base model and lifting its reasoning into 3D with clever multi-view rendering.
Scaling up image tokenizers is challenging because they tend to collapse. This work introduces GigaTok, which is a massive tokenizer with superior reconstruction performance. Decoder scaling and regularization helped with stability and overall quality.
C3PO introduces a new test-time optimization technique that improves accuracy in Mixture-of-Experts LLMs by re-mixing expert weights based on similar reference samples.
Using MCTS to quantify sample difficulty, ThinkLite-VL improves reasoning in VLMs with just 11k training samples and no distillation.
OpenAI's BrowseComp is a new benchmark of 1,266 problems designed to evaluate AI agents' browsing skills in gathering complex, hard-to-locate information online.
Executives from nine companies share how they're leveraging Google Cloud's AI tools to drive innovation across sectors, with over 600 real-world use cases highlighted.
Vertex AI introduces updates to video, image, speech, and music generation models, enhancing creative workflows for businesses. Google AI is enabling specialized AI agents for companies, improving productivity and security. A new Agent2Agent Protocol allows different AI agents to securely communicate across platforms.
DeepSeek's inference engine is built on VLLM, although it is now heavily modified.
NVIDIA is localizing AI hardware production by building factories in Texas and Arizona, aiming to produce Blackwell chips and AI supercomputers entirely within the U.S.
Educators can now use Gemini to generate questions or quizzes from selected text in Google Classroom, enhancing lesson interactivity and streamlining content creation.
Mantra's OM tokens plunged 90% within a few hours, dropping from over $6 to 40 cents. The team claims there were forced liquidations by centralized exchanges, triggering a cascading price drop. Prominent industry figures have condemned the incident, some alleging that the Mantra team sold their tokens OTC and have abandoned the project while others call for more transparent centralized exchange liquidation data.
Kraken has begun rolling out commission-free trading for over 11,000 U.S.-listed stocks and ETFs in select states, aiming to eventually serve all eligible U.S. clients. This move aligns with Kraken's vision of creating a 24/7, all-in-one platform for trading crypto, equities, and other asset classes and sets the stage for future asset tokenization.
Uniswap Wallet now uses "Flashbots Protect" to automatically shield all transactions from front-running and sandwich attacks, ensuring users receive the best swap outcomes such as low slippage.
Privy now supports login via WhatsApp, enabling a smoother and more cost-effective onboarding experience for global users compared to traditional SMS. This feature offers developers an easy path to onboard users and layer in additional security like MFA as wallets grow in value.
The crypto AI agent bubble surged past $20 billion in market cap before rapidly declining to $4-6 billion, signaling a shift from hype-driven speculation to a focus on underlying infrastructure and real-world utility. Early agents were characterized by flashy, speculative demos, but the market is now gravitating toward decentralized AI solutions that allow for data ownership and practical applications. Both web2 and web3 projects are honing in on fundamentals like improved UI/UX, verifiable agentic workflows, and cost-effective deployment.
ERC-7683 is a new Ethereum standard that enables crosschain Intents, letting users perform actions like swaps or votes across Ethereum, L2s, and sidechains in a unified, standardized way. By reducing fragmentation, improving UX, and boosting liquidity and transaction efficiency, ERC-7683 aims to unify Ethereum's ecosystem and drive the next wave of crosschain innovation.